Jump to ContentJump to Main Navigation
Liberalism, Childhood and JusticeEthical Issues in Upbringing$

Tim Fowler

Print publication date: 2020

Print ISBN-13: 9781529201635

Published to Policy Press Scholarship Online: September 2020

DOI: 10.1332/policypress/9781529201635.001.0001

Show Summary Details
Page of

PRINTED FROM POLICY PRESS SCHOLARSHIP ONLINE (www.policypress.universitypressscholarship.com). (c) Copyright Policy Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in PPSO for personal use. Subscriber: null; date: 18 January 2022

The Case against Neutrality

The Case against Neutrality

(p.55) 6 The Case against Neutrality
Liberalism, Childhood and Justice

Timothy Fowler

Timothy Fowler

Policy Press

Abstract and Keywords

In this chapter I consider and reject various leading arguments for liberal neutrality. I show that the extent of so called ‘reasonable’ disagreement about the good is much less than is often supposed. Much of the disagreement about ethics exist because of the human tendency to adopt the beliefs of one’s community. This tendency, rather than the limits of reason, better explains why many views continue to be widely held. Since children have a powerful interest in holding an epistemically reasonable conception of the good, they have an interest in being raised with a sound view of the good. I argue against the view that equal respect requires state neutrality, since respect for children requires giving them a good upbringing. Finally, I show why perfectionism for children need not be illegitimate.

Keywords:   Neutrality, Political Liberalism, Burdens of judgement, Stability

In this chapter my aim is to show why the leading arguments in favour of liberal neutrality fail and are particularly problematic in the case of children. As we saw at the end of Part II, developing a metric of children’s flourishing is hindered if it is impermissible to take a stand on contested ethical questions. Here I argue that there are no good grounds to justify the neutrality restriction. In the ongoing argument these discussions create the space for a perfectionist theory of upbringing. According to such a theory, children are owed an environment conducive to their flourishing according to the best available theory of well-being.


Before beginning the discussion of liberal neutrality it is worth briefly adumbrating the relevant terminology and some core features of the relevant theories. I use the term ‘liberal neutralists’ to refer to any theorist who accepts the principle of neutrality, defined as the view that ‘the justification for laws and policies must not rely on any contested conception of the good’. A ‘conception of the good’ is a set of beliefs and commitments about what human flourishing requires. The most widely discussed liberal neutralists are a group of theorists known as ‘political liberals’, following Rawls’ characterization of this position. Political liberalism is a complex view, but the essential argument is that the state, and all principles of justice that govern state behaviour, must remain neutral because of ‘the fact of reasonable pluralism’. Disagreement about ethics is the natural result of human reason under conditions of freedom, such that in any free society we should expect disagreements about morality or ethics among reasonable people. The word ‘reasonable’ here refers to people who are acting in good faith and who accept that others may disagree with them about ethics. This disagreement justifies neutrality (p.56) because of the liberal principle of legitimacy, according to which the use of public power must be justified to all citizens in terms they can reasonably accept. I explore this principle in more depth later in the chapter, but in brief Rawls believes that citizens can reasonably reject laws premised on ethical beliefs they do not share. The result is that participants in debates on ‘constitutional essentials’ (or perhaps all policies, see Quong, 2004) cannot appeal to their own beliefs about what makes for a good life. Instead, political liberals believe that policies are justified according to an ‘overlapping consensus’ of liberal values to which citizens can all subscribe despite their ethical commitments. Some of my arguments are directed specifically at political liberals and I will make it clear when that is the case; other arguments target the wider set of liberal neutralists who offer other arguments for neutrality. The alternative to liberal neutrality is a family of theories known as ‘perfectionist’. Perfectionists take the view that the role of the state is to enable people to live good lives; thus, knowing what kind of institutions should exist requires a guiding theory of what makes for a good human life.

Arguments for neutrality


Epistemic arguments for neutrality rest on the difficulty of knowing what constitutes a flourishing human life. In Chapter 3 I put forward a list of valuable functionings as constitutive of a flourishing life. The epistemic objector asks how I know, or can prove, that goodness consists of the items on this list. To make sense of this kind of objection, it is useful to distinguish between a strong and a weak variant:

Strong epistemic neutrality: There cannot be ‘knowledge’ about ethics even in principle since all ethical claims are merely statements of feelings or preferences.

Weak epistemic neutrality: There may be truths about ethical matters, but ethical knowledge cannot be held with sufficient confidence to justify perfectionist policies.

The strong epistemic argument for neutrality is a form of moral scepticism, a widely held position in meta-ethics according to which there is no such thing as a ‘moral fact’ and thus no ‘right’ answer to moral questions (Sinnott-Armstrong, 2006). Instead, moral opinions are best understood as mere preferences or tastes. This seems to provide an argument for neutrality, in so far as it leads to the conclusion that there are no good reasons for the state to pursue any ethical value over another. (p.57) There are many different varieties of moral scepticism and assessing their merits is well beyond the purview of this book, but, fortunately, such an assessment is unnecessary because ethical scepticism cannot provide a sound argument for political neutrality. Ethical scepticism not only impugns claims about what constitutes human flourishing but rules out all moral claims, including the positive arguments for neutrality and beliefs about what justice requires. According to liberal neutralists, justice requires that the state takes no stance on matters of ethics or metaphysics. Their view is thus that there are moral reasons to favour one set of policies (the neutral ones) and to disfavour others. But if such scepticism is true then there are no moral reasons of this kind. The neutralist’s view, in accordance with mine, is a substantive view about what morality requires in the case of children. Defenders of neutrality must provide reasons that count against promoting controversial ideals but that do not undermine their own moral ideals.

The modest epistemic argument for neutrality is more promising. Ethics is undoubtedly a complex business and the views of well-informed and sensible people differ substantially. The state should be neutral because it cannot be sure which ethical views are right and which are wrong. The most prominent modest epistemic arguments draw on a concept from Rawls’ Political Liberalism, known as ‘the burdens of judgement’, though I do not think Rawls is really making an epistemic argument.1 The burdens of judgement are a collection of factors that explain the disagreement about ethics among reasonable people. They include the fact that matters of ethics are highly complex, that people’s history can bias them towards some ethical opinions over others and that it is difficult to access all the relevant evidence (Rawls, 1993: 54–8). Rawls believes that while, in the long run, we should expect some consensus among relevant experts on questions of empirical fact, the burdens of judgement imply that people will persistently disagree about even the most basic elements of ethics and theology. While the burdens of judgement are a specifically political liberal idea, other epistemic arguments have also been advanced that draw on further reasons to doubt the strength of ethical knowledge. In various ways, things such as the burdens offer an argument for neutrality in something like the following form:2

Premise 1: The state ought not to promote a conception of the good without sufficient confidence in its soundness.

Premise 2: The burdens of judgement – or a similar idea – mean that no ideal can be held with sufficient confidence to justify its imposition on others.

While each premise of the argument is initially plausible, both must eventually be rejected. The problem with premise 1 is that there are often (p.58) costs to inaction as well as to action. Premise 2 overstates the scope of the burdens of judgement (or other similar considerations); while ethics is complex there are many ideals which can be held with justified confidence.

To illustrate the problem with premise 1, consider the following case:

School choice: A bureaucrat, Martha, is deciding whether to fund a small local school, or to use the money to expand a larger school further away and send all the local children there. Most students at the local school went to nursery together and can be expected to continue their friendships and spend a lot of time in each other’s company outside of school. The other school is further away but is a leading school academically and has access to a greater range of extra-curricular activities such as art or music.

While mundane, Martha’s decision turns on complex issues about what would be best for the children. Keeping the local school open will facilitate stable relationships and will reduce the costs of travelling for both children and their parents. Conversely, sending the children to the larger school promotes learning and extra-curricular attainment. Making this choice therefore requires some heuristic for comparing the importance of artistic achievement against the boredom of commuting, or ease of access to friends against likely income gains from better academic results. These kinds of questions are precisely the ones to which the burdens of judgement should apply. Nevertheless, it cannot be impossible for Martha to make a reasoned choice (and, note, her choice in this matter would be an action of the state).

Now suppose that if Martha does nothing the result will be that both schools are shut down so all the children will end up at a school that is even further away and does not score well academically. The uncertainty between the first two options does not count as a reason against making some choice, because Martha is sure that doing nothing would lead a very bad outcome. In many instances we have reason to think that the costs of inaction are high and thus (absent other arguments) sometimes the state should act even when extremely uncertain about what is best because there may be good reason to think that the status quo is worse.

The problem with premise 2 is that while there are certainly some ethical issues on which it is very difficult to come to an answer, there are many others in which this is not the case. More formally, while the burdens of judgement are a plausible way of accounting for some of the ethical disagreements in society, they are a very poor explanation for some other aspects. Rather, a great deal of the controversy about the good is explained by factors including:

  1. 1. People uncritically following the views of their parents or local communities.

  2. (p.59) 2. The self-interest of dominant groups in society.

  3. 3. A pernicious media environment driven by ideology and with poor fact checking.

  4. 4. Lack of public trust in figures who have access to the right information.

These factors (and many others) are sources of unreasonable disagreement. By this I mean a disagreement in which the views of one side are not held with sufficiently good evidence or rationale and would be rejected by all or most relevant experts. Both sources of disagreement can operate at the same time. It is entirely consistent to accept that burdens of judgement exist and explain some forms of disagreement but still to believe that much ethical disagreement in contemporary societies is unreasonable rather than reasonable in nature.

Dissecting which beliefs are epistemically reasonable and which are not is a daunting task. However, all that I need in order to show why the state should not be neutral about all ethical disputes is to claim that many children come to hold patently unreasonable ethical views, and that holding these unreasonable views is bad for their welfare both directly (because possessing knowledge is itself an aspect of the good life) and indirectly if it leads them to pursue aims which are less valuable than other options. The idea is that the state should at least be permitted to act against the propagation of ethically unreasonable views to children, since children’s malleability makes them vulnerable to accepting things without good reason that turn out to be bad for them and for others. The theoretical implication is that the state should only be neutral on matters on which there is genuine reasonable disagreement caused by the burdens of judgement and that many other ethical views will not count as reasonable.

This seemingly modest suggestion – to limit the scope of reasonable disagreement to views that people could accept with good reasons – goes against the writings of many political liberals. Rawls writes that ‘there are no restrictions or requirements on how religious or secular doctrines are to be expressed; these doctrines need not, for example, be by some standards logically correct, or open to rational appraisal, or evidentially supportable’ (2001b: 592). Quong suggests that, on the political liberal view, ‘we are not meant to require anything of anyone’s comprehensive views, other than that they must be compatible with a liberal conception of political justice’ (2011: 296). The set of things that counts as ‘reasonable’ is thus much larger than the set of beliefs which can be held with good reason. There can be no epistemic argument for neutrality about even these ethical matters, since political liberals like Quong and Rawls have conceded that some supposedly reasonable ethical doctrines are indeed patently false.

(p.60) Therefore, the continuation of a large set of false empirical beliefs about the world and faulty moral beliefs is not best explained by appealing to deep facts about the limits of reason. Instead, much disagreement persists because people, and especially children, are likely to accept the beliefs of the community around them. Therefore, I can fully accept that there are good reasons for principles of justice to avoid taking a side on ethical matters which are genuinely unclear, but there is no reason to think this kind of epistemic argument can support liberal neutrality in all or even most cases. Thus far, defenders of liberal neutrality have offered no reason why justice is not furthered by working against the sources of unreasonable disagreement and creating conditions under which children are likely to come to hold a more plausible conception of the good.

Equal concern and respect

A number of prominent writers have defended liberal neutrality by appealing to what Charles Larmore terms ‘a norm of equal respect’ (1987: 59–60). Ronald Dworkin believes that the state owes its citizens equal concern and that this justifies neutrality because ‘since the citizens of a society differ in their conceptions, the government does not treat them as equals if it prefers one conception to another’ (2000: 191). Dworkin’s idea is that the state cannot favour the aims of any its citizens even if they are in fact better or more worthy and, as such, a perfectionist state seems wrongfully biased in favour of some of its citizens.

However, the equal concern argument reveals the ongoing problem with theories of justice assuming that the proper subject is the adult citizen, who comes complete with an existing conception of the good. Considering only adults, it seems reasonable to say that respecting a person means respecting the beliefs they already have. But in the case of children the situation is very different. As I have repeatedly stressed throughout the book, children’s interest is not in having their beliefs respected (since young children do not have beliefs of this kind), rather it is in acquiring a conception of the good that is right for them and does not cause them to undermine the interests of others. The result is that proper concern for a child does not provide a reason to be neutral about ethics. Instead concern for that person pushes towards ensuring that they are guided to things that are good for them rather than ways of living which will hold them back.

Therefore, the implications of equal respect provide another way in which the case of children must be thought of as central to questions of justice, not an issue that can be tagged on later. Given that everyone is a child during some part of their life, it does not make sense for anyone (p.61) to say that all that respect entails is liberal neutrality. As Quong writes, ‘perfectionism allows the state to treat citizens as if they were children, in need of guidance and direction about their own lives’ (2011: 315). Here Quong seemingly concedes that perfectionism directed at children would be permissible, at least in so far as the duty of respect is concerned. What he misses is the sweeping importance of this, perhaps because the case of children is taken to be a somewhat side issue. Instead, our duties towards children affect not just the design of the education system, and how laws should govern parenting, but rather all our interpersonal conduct that together creates the culture in which children are raised. While it is often disrespectful to try and guide a person once they are an adult, during childhood everyone does need guidance and direction about their own lives. What form that direction should take and what balance come from the state, other citizens or parents is a large question in Part IV. Here though, I only claim that it shows a child no disrespect to design their upbringing in order to guide them towards good ways of living, and in fact we show them disrespect by thinking of them as already beholden to the choices of their parents or the community into which they are born.

Political liberal legitimacy

The final, and most significant, argument for neutrality is that only neutral laws can be properly legitimate. Legitimacy as a concept distinguishes rightful exercise of power from mere force. Often, we think about legitimacy in democratic terms, but political liberals believe that mere majoritarian democratic legitimacy is too weak, since it allows majorities to impose their will on minorities. Instead, following Rawls, they argue that a law is only properly justified if it is justified to every citizen in terms they can reasonably be expected to accept (Rawls, 1993: 136). If any citizen has a reasonable objection to a law, then this law is not properly legitimate. This principle of legitimacy provides a justification for neutrality because in a pluralistic society there are no shared ethical or metaphysical premises which can be used to justify the use of power to all citizens. For instance, a Muslim citizen cannot be expected to accept a law justified by reference to the truth of Christianity. As a result, no laws justified according to the truth of Christianity can be legitimate across a population that includes even a few Muslim citizens (or atheists or adherents of other faiths), even if the majority of the population is Christian and wants a law justified on biblical authority. The same reasoning is taken to apply to laws justified according to any views about ethics or metaphysics, since there will always be some citizens who cannot (p.62) accept these ethical theories. The result is that only neutral laws and policies, supposedly acceptable to all citizens regardless of their ethical creed, can ever be legitimate.

In response, I will suggest that the political principle of legitimacy must be rejected because of what is called the asymmetry problem. The asymmetry in question is the different ways that political liberals respond to disagreements about politics as opposed to those about ethics or theology. When citizens disagree about ethical matters, political liberals argue that it would be morally wrong to favour either side. However, they respond to disagreement about politics very differently, as Rawls, Quong and many others favour redistribution of wealth and increased social spending. These economic policies are highly controversial and their aims and justification would conflict with the sincere convictions of many citizens. Just as they are characterized by deep disagreement about ethics, current liberal societies are characterized by polarization and deep disagreement about politics.3

The asymmetry challenge forces the neutralist to take one of two routes. They must either: (i) show why disagreements about ethics are different from disagreements about politics; or (ii) apply the principle of neutrality to both ethics and politics in a coherent and plausible way. The first strategy is pursued by Rawls himself, but most comprehensively developed by Quong, the second is explored by Gerald Gaus. I will argue that neither route is convincing.

Quong’s solution to the asymmetry problem is to argue that there is an important moral difference between ethics and politics, such that only the first demands neutrality. In Quong’s terminology, disagreements about ethics are always ‘foundational’, whereas political conflict is sometimes ‘justificatory’. A foundational disagreement occurs ‘when the parties do not share any premises which can serve as a mutually acceptable standard of justification’ (Quong, 2011: 193), whereas a justificatory disagreement occurs when ‘participants share premises that serve as a mutually acceptable standard of justification, but they nevertheless disagree about certain substantive conclusions’ (Quong, 2011: 204). Quong argues that it is impermissible for one party to impose a policy based on contested beliefs when other parties foundationally disagree, but that imposition of contested ideas is permissible when the disagreement among the parties is justificatory.

Quong’s distinction provides a solution to the asymmetry problem only in the case of disagreements that may occur among citizens whom he defines as reasonable. In this context, being ‘reasonable’ means sharing a set of basic liberal views about justice; thus such citizens’ disagreements about politics will be justificatory. They all share the same basic political views, they just disagree about how best to apply them. According to (p.63) Quong, it is permissible for reasonable citizens to impose political ideas on each other, since it is allowable to impose contested ideas within justificatory disagreements. By contrast, even among reasonable persons there will still be foundational disagreements about questions of ethics. This shows that in a highly idealized society of reasonable people there is a moral difference between imposing contested political ideas as opposed to contested ethical ideas.

However, I believe that Quong’s solution, while elegant, faces a dilemma. Either he will draw the category of reasonable citizens too broadly, in which case the asymmetry problem re-emerges between reasonable persons, or he will draw it too narrowly, in case we cannot consistently assume everyone will be reasonable even in an idealized well-ordered society. A broad understanding of reasonableness is suggested when Quong (2010: 233) argues that reasonable citizens need only endorse basic liberal values like freedom and equality. On this broad understanding, a citizen could be reasonable and hold a wide array of economic and political views. The problem is that contemporary societies have shown us that people can agree that freedom and equality are valuable yet continue to disagree, fundamentally, about political matters. For instance, some people believe that governments which intervene in the marketplace necessarily undermine freedom and individual rights. These people do not see the creation of welfare states as merely a slightly different interpretation of a principle they already hold. Rather, they see such states as deep threats to their core ethical beliefs, which violate important rights that for them are just as important as the social rights that institutions like universal education or health care are meant to protect. If such conservative citizens can count as reasonable then the asymmetry problem remains a live issue even in the ideal case. It is unclear how welfare states can be legitimately imposed onto them.

Alternatively, the category of reasonable citizens could be constituted so that these people are united not just by a basic commitment to freedom and equality but by a much thicker set of commitments. This is the narrow view about what being a ‘reasonable citizen’ means. On the narrow view, by definition, to say that a person is reasonable is to say that he or she accepts that justice requires social spending and progressive taxation. Disagreements between narrowly reasonable citizens are more similar to what might occur within a very ideologically coherent political party, and such disagreements do indeed look justificatory. However, while the narrow reading does imply that all disagreements about justice among reasonable people are foundational it creates different but equally serious problems.

For a start, the narrow reading radically redefines who counts as ‘unreasonable’ in ways that do not match how this term is often used (p.64) or understood. For instance, Quong and other political liberals often write as if the category of unreasonable citizens is synonymous with that of dangerous extremists. We are told that ‘Unreasonable doctrines are thus doctrines whose beliefs directly contradict the fundamental political values of a liberal democratic regime’ (Quong, 2010: 302). Examples offered of unreasonable citizens include misogynists and members of the KKK. However, on the narrow reading, Quong is not entitled to talk about the class of the unreasonable in this way. Racists or misogynists are indeed unreasonable by this definition, but so are those who believe their religious views should have any political consequences or who believe that state action in the economic sphere is a threat to personal freedom. In short, a very large proportion of people in modern society will count as unreasonable on this narrow reading.

The worry is that while Quong can show why it is acceptable to impose progressive tax laws onto the set of people who already agree with them, he has not shown why such people are justified imposing their views on anyone else except to say that these others are unreasonable. But saying that someone is unreasonable does not seem like a good reason to discount their view, once it is granted that the class of unreasonable people is so large. Therefore, neither the strong nor the broad understandings of what reasonable might mean ultimately show why it is permissible to impose left liberal views of justice, but not sometimes permissible to impose views based on a contested conception of the good.

The second kind of response to asymmetry is explored by Gerald Gaus, who argues that the state cannot legitimately impose polices justified by any sectarian political or ethical beliefs. In his earlier work, Gaus (2003) argued that this implied that almost no actions by the state could ever be legitimate. In his more recent work, Gaus and others have explored various ways in which some kinds of social structures might be legitimate despite deep disagreement about justice (see Gaus, 2016; Vallier, 2014). In brief, Gaus believes that a relatively minimal state could be legitimate because it performs functions upon which everyone would agree. Within a minimal state people are free to band together to create social institutions, but they cannot impose these structures on everyone since some will disagree with them. There are some serious concerns as to whether Gaus’ solution really is legitimate by his own lights, but the larger problem is that he makes the price of accepting the principle of political liberal legitimacy too high, especially in the context of past or current injustices.

In the economic sphere, a serious problem with Gaus’ minimal state solution is that it permits previous injustices to continue to have long and significant effects. Consider the following case:

(p.65) Status quo: A small group of people become wealthy via the exploitation or persecution of other groups. This wealthy elite hoard their wealth and privileges and pass them only to their own children. These children come to hold political views by which they are entitled to this wealth and status (these children are not duplicitous; they genuinely believe they are entitled to their wealth).

According to Gaus, the mere fact that these advantaged people hold the belief that they are entitled to their wealth rules out redistributive policies. The rich object to redistribution and even a democratic majority does not have the right to impose views of justice over conscientious objections. The earlier injustice thus has persisting effects because there is no legitimate way to rectify it. A similar kind of pernicious status quo can persist with respect to the ethical beliefs that children will come to hold. Mistakes about ethical matters made by previous generations have continuing effects because the mere fact that people hold mistaken beliefs is enough to block actions to combat these beliefs (even including giving children an education that allows them to critically dissect these beliefs, since some parents conscientiously object to their children receiving an advanced education).

Therefore, I believe that both potential solutions to the asymmetry problem fail, for deep rather than easily solvable reasons. The implication is that the political principle of legitimacy must be rejected. Rather, legitimacy should be understood in some other way; perhaps simply as pertaining to laws that pass a suitably democratic procedure (see Waldron 1999: 282–94 and Christiano, 1996: 35), or that legitimate laws are ones that best serve the public good. Either of these alternatives allow that sometimes perfectionist actions can be permissible and, as such, do not support liberal neutrality.


In this chapter I surveyed the literature on liberal neutrality, which is a necessary enquiry given this principle rules out many elements of my understanding of children’s justice. I argued that none of the arguments for liberal neutrality provide good reasons to object to perfectionism, especially in the case of children. I highlighted that the principle of neutrality is at its most plausible when we consider adult subjects with well-formed beliefs and the ability to control who they interact with. In this case, a principle of letting each person get on with their own plans seems plausible and respectful. However, the case for neutrality unravelled when applied to children, as they are profoundly vulnerable to the ethical beliefs of others.


(1) For a presentation of an epistemic case for political liberalism, see Peter (2013).

(2) For defences of neutrality, albeit not by political liberals, see Barry (1996: 168–73) and Lloyd-Thomas (1998: 115–25).

(3) For instance, see Achen and Bartels (2016) for a discussion about the importance of political polarization in the formation of tribal identities.